Studies of the Lumbar Vertebral End-plate Region in the Pig

The vertebral end-plates and the epiphyses from the lumbar region of the spine in juvenile pigs were studied. Cell density of the hyaline cartilage of the vertebral bodies was measured as well as cartilage thickness. The relative bone content in the region of the nucleus pulposus was calculated. A certain negative interrelationship between cartilage thickness and cell density of the hyaline cartilage of the vertebral bodies was found. The relative bone content of the cranial epiphysis was higher than that of the caudal epiphysis in the same motion segment at all levels of the lumbar spine. The difference was pronounced down to 2 mm from the cartilagelbone interface


INTRODUCTION
The vertebral body consists ofa core of cancellous bone surrounded by a thin layer of cortical bone ( 3 ) .In each vertebral body there are also the two epiphyseal growth plates and the epiphyses.The epiphyses consist of cancellous bone distally covered by hyaline articular cartilage; the end-plates, forming an interface between the bone and the intervertebral discs.
The cancellous bone of the epiphysis is vascularized by capillaries and the blood reaches the bonelcartilage interface through a rich sinusoidal network ( 3 . 4 ) .It has previously been demonstrated that the central area of the distal part of the epiphysis (adjacent to the nucleus pulposus and the inner annulus fibrosus of the disc) of the human adult lumbar vertebrae is supplied with nutrients via the nutrient and metaphyseal arteries, whereas the peripheral area of the distal part of the epiphysis (adjacent to the outer ring of the annulus fibrosus) is supplied via the peripheral arteries (18).It has also been suggested that the peripheral arteries are less subject to arterial degeneration than the nutrient and metaphyseal arteries (18).Brookes (1977) stated that the arterial blood flow to the vertebrae and the vascularisation of the entire vertebral body is rich, which suggests that the oxygen tension of the blood here is relatively high.Brookes concluded, that this high oxygen tension was probably linked to the haematopoietic function of the bone-marrow of the vertebrae and the dominating trabecular architecture of the bone within the vertebral bodies ( 3 ) .The vascularisation of the vertebral bodies has also been qualitatively described by others (4,5,8,18,20,21).
The intervertebral discs are situated between the vertebral bodies.The disc is the largest avascular tissue of the body and is for its nutrition completely dependent upon nutrients entering from surrounding blood pools.
It has previously been demonstrated that there are two main nutritional routes into the intervertebral disc, i.e. transport a) from the blood vessels surrounding the periphery of the annulus fibrosus, and b) through the end-plates (predominantly in the region of the nucleus pulposu~)(2,6,11,16,18).
The nutritional status of the cells of the central nucleus pulposus is precarious; the oxygen tension field as well as the concentration gradient of glucose is very steep and the cells in the central region of the tissue may not be able to satisfy their requirements (10).Cell necrosis in the disc due to lack of substrates such as oxygen and glucose may lead to severe disc degeneration.It has been suggested that one reason for changes in the discal nourishment may be changes in the vertebral body blood supply (18).
Certain features of the epiphysis govern factors that are of special importance for the nutritional status of the central part of the disc: 1) The arterial blood that reaches the vertebrae enters the blood pools from the sinusoidal network of the epiphysis.Nutrients diffuse into the marrowspaces and extra cellular liquid in the cancellous bone of the epiphysis and further from these spaces into the disc.The amount of blood pools and marrow-spaces in the immediate vicinity of the bonelcartilage interface is therefore of importance to the transport of solutes to the cells of the disc.The route of diffusion via the peripheral part of the annulus may for certain solutes be too long to contribute to any considerable extent.
2) Nutrients must, in order to reach the disc, diffuse through the hyaline cartilage of the end-plate.The cell density of this cartilage is much higher than that of the central part of the disc.This relation has been established to be approximately 2.8:1 (10).A certain amount of nutrients leaving the bone will therefore be consumed by the chondrocytes of the hyaline cartilage.
Hence, the thickness and cell density of this cartilage is of importance to the nutritional status of the cells of the disc.Ingelmark and Ekholm (1948) reported that the thickness of the hyaline cartilage of the knee joint in the rabbit varies with physical exercise.They observed a 10-12% increase in cartilage thickness after 10 minutes of dynamic exercise (running), compared to the cartilage thickness after 60 minutes of rest in a position where the knee joint was not stressed.It was stated that this rapid increase in cartilage thickness was due to an increase in fluid content of the cartilage (12), and that flow of fluid from the marrow cavities gave higher controbution than from the articular space (12).No such increase in cartilage thickness with physical exercise has been reported concerning the hyaline cartilage of the vertebral bodies.Bernick and Caillet (1982) observed age changes in the arterioles, capillaries and venules in the nutrient canals or spaces of the bone adjacent to the cartilage or disc in human lumbar vertebrae.They also observed that the articular cartilage of the end-plate undergoes calcification followed by resorption and replacement by bone with increasing age.They stated that this calcification of the articular cartilage and the vascular changes seen in the older vertebrae would inpede the passage of nutrients from the blood to the disc (1).
3) Since one of the main transport mechanisms of solutes to the cells of the disc is diffusion through the end-plates,the distance between the vertebral bodyldisc interface and the central disc is of importance.
The aim of this investigation was to study the above mentioned features in the end-plate region.

MATERIALS AND METHODS
Seven juvenile pigs of both sexes (9-10 months old, average weight 63 kg) were included in the investigation.Initially, an intramuscular injection of Ketalar (Parke-Davis, Detroit, Mich., USA) was administered.After that a catheter was inserted into a vein of the ear an?. an overdose of Pentothal sodium (Abbot Laboratories, Chicago, Ill., USA) with KC1 was administered to kill the an ima 1 .
The entire lumbar spine together with the major part of the thoracic spine was excised, quickly freed of soft tissue, frozenand subsequently stored - The spine was later, while being kept frozen, sectioned by an electrically powered bandsaw into vertebral motion segments, each motion segment consisting of one intervertebral disc and half the vertebral bodies immediately cranially and caudally to it, (Figure 1).The motion segments were then cut in half by a medial sagittal cut.each decalcified slice (Fig. 2).This piece was embedded in paraffin wax and sectioned further into 5 pm thin slices on a sliding microtome.The slices were then strained in haematoxylinleosin ( 9 , 1 4 ) .The interface between bone and hyaline cartilage is not straight and sharply defined due to the rich vascular sinusoidal network of this area (Fig. 3 ) .Here an arbitary line had to be drawn.However, the measurements were at a fairly great magnification which simplified the determination of the position of the interface.The error was approximately -+lo%.
Even at a considerable magnification it is sometimes difficult to decide whether what is seen under the microscope is actually a chondrocyte, just an empty lacunae or possibly two cells in the same lacunae.However, as a large number of measurements were made, the magnitude of this kind of error was minimized to approximately 21 5%.

Measurements of bone content.
The estimation of the area and the subsequent weighing procedures were done manually.In this method there are several sources of possible errors.The resolution at a magnification of x50 is good, but not absolute.The ink of the pen with which the marrow-space were drawn on the plastic sheet has a certain weight which has been neglected.However, the method is simple, but nontheless quite accurate.The entire procedure of determining bone content was repeatedly performed and the reproducibility was found to be acceptable.The error averaged 52%.

Macroscopic dimensions of the intervertebral discs of the pig
The anterior-posterior width of the disc varies between different levels of the lumbar spine.In this study the mean disc width at the Thl5/L1 level was 18.56 mm.The width then gradually increased at the LlIL2 and L2/L3 levels.
The highest mean value, 21.05 mm, was obtained at the L3/L4 level.A further descent of the lumbar spine showed a decreasing disc width and the lowest mean value of the entire lumbar spine, 17.78 mm, was found at the L5/L6 level (Table 2).
The caudal-cranial distance from the cartilageldisc interface to the central disc did not vary significantly between different levels of the lumbar spine.However, our results indicated that the central part of the disc was somewhat thicker at the Thl5/Ll and L5/L6 levels, 1.92 rn and 2.19 respectively (Table 1 ) .The difference in bone content between the two epiphyses of one motion segment was measured in two different samples, both at the L3/L4 level.
This difference was observed also in all other sections, from different levels of the lumbar spine.When studying a motion segment under the micro-, scope it could easily be determined (from the sole criterion of bone content), which epiphyses was the caudal one and which was the cranial one.The same observations were made when studying sections of motion segments from dog (unpublished results).
It has previously been suggested that bone mineral content determined by dual photon absorptiometry is higher in the cranial than in the caudal part of one motion segment in man (7).It is also an impression that healing microfractues in human vertebrae predominantly have been found in the caudal epiphysis of a motion segment (7).These circumstances together with our finding: of a lower relative bone content in the caudal than in the cranial epiphysis of one motion segment indicate that the caudal epiphysis is more sensitive to the effects of demineralization than the cranial epiphysis.This would be the case already in a juvenile and healthy individual.
The difference in relative bone content was observed to a depth of 1.5 nnn measured from the articular cartilage.At greater depths there was no noticable difference.However, blood vessels and blood pools at greater depths than approximately 1.5 ma are probably of little importance as far as diffusion of nutrients to the disc is concerned.
A lower relative bone content in the caudal epiphysis of a motion segment obviously means a higher relative content of marrow spaces.These spaces contain the blood vessels and blood pools on which the disc is highly dependent for its nourishment.A microfracture in this epiphysis might therefore have consequences not only for the vertebra itself, but also impair the transport of nutrients tothe intervertebral disc, something which might be deleterious to the cells of the nucleus pulposus.

Fig. 1 .
Fig. 1.Schematic drawing of a motion segment consisting of one intervertebral disc and half the vertebral bodies immediately cranially and caudally to it.The drawing shows sections indicated as A -A and al-a4: strips of the epiphy-1 4 sis (0.5 mm wide) where relative bone content was measured.B -B and b -b : sites in the hyaline end-plate cartilage where cell density and cartilaAe zhickness were measured.The symbols indicate sites adjacent to the following regions of the intervertebral disc: B 1 and b * Anterior inner annulus fibrosus.B and bir Central nucleus pulposus.B B5 and b5: Posterior inner annulus fibrosus.D 1 -D3: Posterior peripheral part of the nucleus pulposus.Sites where disc thickness was measured.The dimensions of the discs were measured macroscopically (Fig: 1: D,, D2, D3, and Fig. 2: E, F, G) and each half motion segment was thereafter cut sagittally into 2-3 mm thick slices.During this entire procedure the specimens were kept frozen by dry ice.Some of the slices were fixed in 10% buffered formalin (15,17) for 5-7 days after which they were left for 3-5 days in formic acid for decalcification (17).A rectangular piece with the approximate dimension of 10 x 15 mm was cut out around the nucleus pulposus of the disc in

plateFig. 2 .Methodological problems and errors 1 .
Fig. 2. Vertebral motion segment.E, F, and G show measured anterior-posterior widths of the intervertebral disc.The dashed line indicates the area around the nucleus pulposus cut out of the decalcified slices for further histological studies.

Fig. 3 .
Fig. 3 .Part of vertebral motion segment showing posterior peripheral nucleus pulposus and inner annulus fibrosus, hyaline end-plate cartilage and the cancellous bone of the L3 epiphysis with the uneven bone-cartilage interface due to the sinusoidal vascular network of the vertebra.(Photo-microscope x50).

2 Fig. 5a .
Fig. 4a.Section ofthe L3 vertebral body from the area adjacent to the central nucleus pulposus.Notice the dence cancellous bone, with small marrow-spaces, of the epiphysis.(Photo-microscope x50).

Fig. 6 .DISCUSSION
Fig.6.Relative bone content of the epiphysis in the region adjacent to the nucleus pulposus at the L31L4 level, expressed as a percentage of the total sagittal area.

Table 1 .
Average distance(mu)from the surface of the end-plate cartilage to the center of the disc at different levels of the lumbar spine.The sites of measurement indicated as D1, D and D are shown in Figure1.

Table 2 .
The anterior-posterior width (mm) of